• I have very sensitive amps, 150Watts for 0.75Volts input. I ran attenuators at the amplifier end of the leads
consisting of a series 10k and about a 1k shunt to earth. I used Vishay bulk foils so as to do the minimum damage to the signal path. Not sure now what dB cut that is. I'd probably use 5k now. Don't know why I used 10k, probably already had them for something else. For neatness and ease, you could just as well fit them inside the amplifier case. Or maybe even to the amp input; possibly fit the 5k series and cut the (often 47k) resistor to earth to some like a 500Ohms? Don't know if there's any reason why that is not a good idea. It would change the relationship with an input capacitor to earth.
Having the resistors at the source end of the lead can attenuate the bass and treble.
• Now, I'm changing the gain of the op-amp. Planning to change the two first series resistors from 2k7 to 3k3. Mainly because I have them already, Vishays again
If the maths are correct, that lowers the gain from about 3.5 to about 3. However, if you use 4k7 that lowers it to about 2.4.
You'll also need to change the capacitor that follows the two first resistors. 434pF goes with 4k7, as best as I can work out anyway. Farnell do a 430pF FSCEX extended foil polystyrene capacitor.
However I just bought 680pF for the 3k3. A shade high really, 620pF equates to the filter slope that Auzentech used.
I'm doing it this way as I want to delete the surface mount parts an already have the Vishays in near values. Another way to drop the gain would be to just change the feedback and balancing SM resistors. They are 6k8. You could go to 2k7 I guess to really cut it to no gain; just the normal 2Vrms output as far as I can make out from the data sheet, or chose a value somewhere in between. You might want to match the resistance value of the resistors you use.
• Option 2 carries the risk of destroying the DAC from an over current condition, eg, short circuit. Output resistors might be sufficient, don't know what value, the data sheet may be helpful for that. The leads will probably pick up quite a bit of RFI and you might need to do better filtering. Maybe use an instrumentation style balanced line receiver as they can cut the common mode better. Or take the filtered balanced lines to XLRs for a balanced output if you're driving a long lead to your amp(s). Or be more committed, dispense with the connectors and solder the leads at both ends
I'm planing to fit Vishays for all the resistors and FSCEX for the caps. Much of it is not too hard and can go under the op-amp on the solder side of the board around and under the 0.33uF // 22nF. Four resistors have to go on top beside the op-amp, two on each side, which will be tight but I think do-able. Four resistors and the two 680pF cap will go where the 4 DC blockers once were. Also tricky. we damaged some of the solder pads when taking off those OE caps.
This is phase one. I can then run the card as it is and see, with it's standard output tracks inc the tiny RLC filter parts and what it sounds like and how much I get from the Vishays.
Then onto phase two which is the fit, onto the solder side, a pair of LME49600 buffers, ostensibly to provide a low output impedance to drive my headphones. This may or may not be an improvement with respect to driving the line to the amplifiers. It is another chip in the signal path, all be it a reasonably good one, but I'll be able to miss out that 200 Ohm output resistor which I found in the past is sonically a good move. There might be a better line driver, maybe a discrete circuit; I'm sifting DIY Audio.com for info.
I have a hand written schematic I reverse engineered from the board. I took off some of the surface mount components to measure their value. Yesterday I sketched out the track layout so I could work out how to fit the much much larger components in place. It's not in a form that I can upload yet, but I intend to draw it out better in case it's proves useful to anyone